influence is positive or negative with respect to climate sensitivity. Climate sensitivity, recall, is the relationship between the change in the global concentration of greenhouse gases (given in units of CO2-equivalent impacts on radiative forcings) and the change in the annual mean surface air temperature (see Chapter Four). If the Earth were a simple system, free of feedbacks and other non-linearly interacting processes, this sensitivity would be a straightforwardly linear one: each doubling of CO2-e concentration would result in an increase of ~.30 ${\displaystyle {\tfrac {K}{W/m^{2}}}}$, which would correspond to a mean surface temperature change of 1.2 degrees C at equilibrium[1].

Unfortunately for climate modelers, things are not so simple. The net change in average surface air temperature following a CO2-e concentration doubling in the atmosphere also depends on (for instance) how the change in radiative forcing that doubling causes impacts the global albedo. The change in the global albedo, in turn, impacts the climate sensitivity by altering the relationship between radiative flux and surface air temperature.

Just as with albedo, we can (following Roe & Baker [2007]) introduce a single parameter φ such that the net influence of feedbacks on the equation describing climate sensitivity:

 ${\displaystyle {\tfrac {dT}{dt}}=\varphi ({\tfrac {dR}{dt}})}$ 5(n)

In a feedback-free climate system, we can parameterize 5(n) such that ${\displaystyle \varphi =1}$, and such that ${\displaystyle \varphi _{0}=\varphi _{t}}$. That is, we can assume that the net impact of positive and negative feedbacks on the total radiative flux is both constant and non-existent. However, just as with albedo, observations suggest that this simplification is inaccurate; ${\displaystyle \varphi _{0}\neq \varphi _{t}}$. Discerning the value of ${\displaystyle \varphi }$ is one of the

1. Roe & Baker (2007), p. 630

185